CN101675614A - Method for synchronizing a clock of a network component with a clock of further network component and network component therefor - Google Patents
Method for synchronizing a clock of a network component with a clock of further network component and network component therefor Download PDFInfo
- Publication number
- CN101675614A CN101675614A CN200880013892A CN200880013892A CN101675614A CN 101675614 A CN101675614 A CN 101675614A CN 200880013892 A CN200880013892 A CN 200880013892A CN 200880013892 A CN200880013892 A CN 200880013892A CN 101675614 A CN101675614 A CN 101675614A
- Authority
- CN
- China
- Prior art keywords
- networking component
- transition delay
- time
- clock
- request
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04J—MULTIPLEX COMMUNICATION
- H04J3/00—Time-division multiplex systems
- H04J3/02—Details
- H04J3/06—Synchronising arrangements
- H04J3/0635—Clock or time synchronisation in a network
- H04J3/0638—Clock or time synchronisation among nodes; Internode synchronisation
- H04J3/0658—Clock or time synchronisation among packet nodes
- H04J3/0661—Clock or time synchronisation among packet nodes using timestamps
- H04J3/0667—Bidirectional timestamps, e.g. NTP or PTP for compensation of clock drift and for compensation of propagation delays
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L7/00—Arrangements for synchronising receiver with transmitter
- H04L7/02—Speed or phase control by the received code signals, the signals containing no special synchronisation information
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Synchronisation In Digital Transmission Systems (AREA)
Abstract
The invention relates to a method for synchronizing a clock of a network component (A) with a clock of a further network component (B) communicatively connected to the network component (A) in a network and a network component, whereby the method comprises the following steps: determining a set of forward transition delays for transmission of messages from the network component (A) to the furthernetwork component (B) and a set of corresponding backward transition delays for transmission of messages from the further network component (B) to the network component (A), selecting a minimum forward transition delay (minF) from the set of forward transition delays, selecting a minimum backward transition delay from the set of backward transition delays (minB), deriving an estimated value for anoffset between the clock of the network component (A) and the clock of the further network component (B) from the selected minimum forward transition delay (minF) and minimum backward transition delay (minB), and adjusting the clock of the network component (A) in accordance with the estimated value for the offset estimate.
Description
Technical field
The present invention relates to method and a kind of networking component of the clock synchronization of a kind of clock that makes networking component and other networking component.
Background technology
In a lot of Distributed Calculation were used, more specifically for the real-time application in the packet switching network, it was very important that information and fiducial time (for example producing the time of this information) accurately are complementary.For example, in packet-based network, can produce data in real time by sending networking component, so that this transfer of data is used for further acquisition to receiving networking component.For this reason, data acquisition system is become grouping, each grouping comprises partial data, and the reception networking component is submitted in these groupings, in receiving networking component data division is lumped together.
Because the variation of transmission path and queuing delay (queuing hold-up), the transmission delay that different grouping experienced can change significantly in the uncertainty mode.In extreme case, early the grouping that sends may be than the more late arrival of the grouping of later transmission.To receive networking component and can correctly handle grouping in order to make, can stab each additional period of dividing into groups, this timestamp is corresponding with the fiducial time of packaged data division.In this case, receiving networking component can make data division accurately be complementary with corresponding fiducial time.
For above-mentioned example, making and sending networking component is necessary with the clock synchronization that receives networking component.Put it briefly, the split-second precision measurement is the crucial prerequisite that enables to develop any high-performance distributed computing application.According to application, must keep accurate absolute time benchmark (for example, the zebra time of coordination or UTC time) or relative time benchmark.
In addition, the flow control of host-host protocol and congested control greatly depend on send networking component (or transmitter) and the measured two-way time (RTT) that receives the packet that exchanges between the networking component (or receiver) and/or transit time (one-way latency, OWD).Information fiducial time of timestamp form spends a certain constant or variable time quantum owing to for example supply, so the estimation of this transmission of Information duration is the subject matter of time synchronized on the packet switching network between transmitter and receiver.In order to handle this problem, at D.L.Mill, " TheNetwork Time Protocol (Version 3) Specification; Implementation andAnaylsis ", RFC 1305, IEFT, at length disclosed so-called NTP (Network Time Protocol) (NTP) has been designed to make on the packet switching network local clock of networking component synchronous among the March 1992.
The core component of NTP is based on the clock offset calculation routine of so-called Cristian algorithm (CA), below will summarize this Cristian algorithm.The request networking component is at transmitting time T
SComprise the time request in interior packet to answering the networking component transmission.Answer the fast as far as possible answer of networking component and will be encapsulated in the time of reception of answering in the grouping and stab T
RBeam back, this answer is grouped in answers time of reception T
RRBe received.According to the following calculating forward transition delay of these time values (or one-way latency of forward direction) (OWD
f) and back to transition delay (or the back is to one-way latency of direction) (OWD
b): OWD
f=T
R-T
SAnd OWD
b=T
RR-T
RIn addition, suppose that the actual one-way latency on this both direction equates, make OWD
fWith OWD
bBetween difference only give the credit to the request networking component clock with the answer networking component clock between side-play amount.This synchronisation offset is calculated as (OWD
f-OWD
b)/2.
The Cristial algorithm has significant disadvantage: only equal actual back during to transition delay when actual forward transition delay, the estimation of synchronisation offset is only accurately.That is, when the communication between two networking components be the symmetry the time.Yet in the situation of asymmetric delays occurring, this calculating will be wrong.In live network, because asymmetrical network topology (postponing such as the random packet in asymmetrical digital subscriber line road (adsl line) and asymmetric route and/or the networking component), one-way latency is normally asymmetrical.Herein, for real-time application, the error key that the systematic error that is caused by the asymmetrical network topology is not caused by the random packet delay, wherein this random packet postpones to be caused by variable one-way latency, for example causes owing to variable queueing delay.
Summary of the invention
Thereby the purpose of this invention is to provide a kind of method and networking component, make it possible to achieve the more precise synchronization between the clock of heterogeneous networks assembly in network.
Utilization make according to independent claims 1 a kind of the clock of networking component and other networking component clock synchronization method and satisfy this purpose according to the networking component of independent claims 8.
According to an aspect of the present invention, clock that a kind of method the makes networking component method with the clock synchronization of the other networking component that communicates to connect with this networking component is provided, described method comprises the following steps: to be identified for from the forward transition delay set of described networking component to described other networking component message transfer, and is used for gathering to transition delay to the independently back of described networking component message transfer from described other networking component; From the set of forward direction transition delay, select minimum forward transition delay; To the transition delay set, select minimum back to transition delay from the back; To transition delay, derive the estimated value of the side-play amount between the clock of the clock of described networking component and described other networking component from selected minimum forward transition delay and minimum back; And the clock of adjusting described networking component according to the estimated value that side-play amount is estimated.
According to another aspect of the present invention, provide a kind of networking component, having comprised: clock; Communicator, be configured to provide with network in being connected of at least one other networking component; Request unit is connected with described communicator, is configured to generate request and sends described request to described other networking component; Receiving system is connected with described communicator, is configured to receive answer to described request from described other networking component; Analytical equipment is connected with described clock with described receiving system, be configured to analyze described reception answer, determine transition delay set and the estimation of calculating the clock synchronization side-play amount; Memory is connected with described analytical equipment, is configured to store the transition delay set; And the clock adjusting device, be connected with described analytical equipment, be configured to estimate to adjust described clock according to institute's offset calculated.
An advantage with respect to prior art of the present invention is: carry out the repeatedly measurement of transit time is gathered to obtain transit time, select minimum value from this transit time set.Therefore, can minimize the error that side-play amount that the variation owing to independent measurement causes is estimated.This change may be owing to be used to make the variation of message by the transmission path of network to cause, perhaps this variation may be owing to for example in the input and output queue of networking component the different retention time of message cause.Another notable feature is that the minimum transition delays of forward and backward direction is to select independently from their set separately.This means minimum forward transition delay not necessarily with minimum after to the corresponding forward transition delay of transition delay.This is because the message of experience minimum forward transition delay not necessarily also experiences minimum back to transition delay usually.Therefore, the advantage that has of the present invention is: the estimation that can calculate side-play amount with error boundary more closely.
Refer to forward transition delay for the back to transition delay and the discussion of forward transition delay accordingly and belong to the same communication cycle that comprises request message and answer message to transition delay with the back.For example, if be sent to the message experience particular forward transition delay of networking component, then the transition delay at the answer message experience of this message is that corresponding back is to transition delay.In addition, being used for processing messages and generating the processing time of answering in the hypothetical network assembly is negligible, then forward transition delay can be obtained two-way time with corresponding back mutually to transition delay.
In advantageous embodiment of the present invention, determine that forward transition delay set and corresponding back also comprise step to the described step that transition delay is gathered: be identified for continuous two-way time of between described networking component and described other networking component, communicating by letter, come down to corresponding forward transition delay and corresponding back wherein said two-way time to the transition delay sum, comprise pre-adopted number up to finding, continuously the two-way time of the two-way time of Que Dinging in interior stable set two-way time makes described set two-way time with depart from less than the maximum deviation of being scheduled to the minimum two-way time in described two-way time of the set till, corresponding forward transition delay two-way time in forward transition delay set in login and set two-way time, and in the back in the transition delay set in login and set two-way time two-way time corresponding back to transition delay.
Thereby determine minimum two-way time and the stability region around this minimum two-way time.Obtaining the stability region has guaranteed to utilize corresponding forward direction and back calculating to transition delay to have more other side-play amount estimation of high accuracy level.By selecting enough little predetermined maximum deviation, can determine two-way time and corresponding forward direction and the back influence that is not subjected to the possible basic change of two communications between the networking component to transition delay in the stability region.For example, in the situation of packet-based communication, predetermined maximum deviation can be chosen as several milliseconds magnitude, this is much smaller than owing to fill or transition delay that the flush packet formation is caused changes (this variation can be 10 to 30 milliseconds a magnitude), and even less than because communication path changes the transition delay variation (this variation can be the magnitude of a few tens of milliseconds) that causes.
In another exploitation of the present invention, determine that forward transition delay and corresponding back comprise to the step of transition delay: send request from described networking component to described other networking component at request time, send answer to described request from described other networking component to described networking component in the time for replying, receive described answer at the described networking component of answer time of reception, calculate forward transition delay by from the described time for replying, deducting the described request time, and calculate corresponding back to transition delay by from described answer time of reception, deducting the described time for replying.Thereby provide a kind of definite forward direction and back straightforward procedure to transition delay.
Favourable improvement of the present invention also is included in the step that encapsulates timestamp from described other networking component to each answer that described networking component sends, and described timestamp comprises the time for replying.Timestamp is to be provided for the further simple and effective instrument of the time for replying of calculating to networking component.This timestamp can be the main information in answering.Alternatively, answer can comprise out of Memory or data and the time for replying that networking component is asked.
Advantageous embodiment of the present invention also comprises the step that sends described request with equal time at interval.In favourable exploitation of the present invention, described request and described answer send as packet.
In another embodiment of the present invention, upgrading forward transition delay set and back continuously gathers to transition delay.Upgrading set herein can comprise: abandon whole set and redefine the new set with new clauses and subclauses.Alternatively, renewal process can comprise: when carrying out new measurement, with the new value of determining of interpolation in set of the continuous time interval.
Description of drawings
Below, will utilize advantageous embodiment to describe the present invention with reference to the accompanying drawings.In the accompanying drawing:
Fig. 1 shows the sequential chart that time delay that the grouping that sends between network node A and the B is experienced is described;
Fig. 2 shows another sequential chart that time delay that the grouping that sends between network node A and the B is experienced is described;
Fig. 3 shows the sequential chart of a series of groupings that send between network node A and B;
Fig. 4 shows the figure identical with Fig. 3, has wherein described other characteristic;
Fig. 5 shows the figure of the measurement of round trip time that uses distinct methods;
Fig. 6 shows the figure of the synchronisation offset estimation that calculates based on measurement shown in Figure 5;
Fig. 7 shows the figure of the synchronisation offset estimation of calculating based on the meter of the stability region of two-way time;
Fig. 8 and 9 shows the comparison diagram of the error boundary of the synchronisation offset estimation of using three kinds of distinct methods calculating; And
Figure 10 shows the figure of statistical error bound.
Embodiment
Below, the known method of the side-play amount between the clock of the clock of a kind of estimation network assembly (network node A or node A) and other networking component (network node B or Node B) is at first described.This method generally is called the Cristian algorithm.Node A is connected on network is connected with Node B and for example communicates via packet in internet environment.
Fig. 1 shows the sequential chart of the packet that sends between node A and B.Has time variable t
ATimeline belong to the clock of node A, have time variable t simultaneously
BAnother timeline belong to the clock of Node B.The arrow that points to another timeline from a timeline has been indicated the transmitted in packets from a node to another node.As shown in Figure 1, node A is in time T
S (A)Send the packet that comprises the time request to Node B.Node B is in time T
R1 (B)Receive this packet.After handling this request grouping, Node B gets quick answer and at time for replying T
R2 (B)To comprise time of reception T
R1 (B)In addition the time of reception of mark stabs and sends it back transmitter node A in interior packet.In time T
RR (A)Node A receives the packet with this timestamp.T
R1 (B)And T
R2 (B)Between time difference be on the Node B processing time and usually than time T
RR (A)And T
S (A)Between time difference want much shorter, and be negligible.Therefore can suppose T
R1 (B)=T
R2 (B)=T
R (B)Figure 2 illustrates this relation.
The time of noticing institute's reference is to be measured by the corresponding clock of node A and B.That is, the clock by node A comes Measuring Time T
RRAnd T
S, come Measuring Time T by the clock of Node B
R (B)Usually, because the side-play amount and the deflection (skew) of corresponding clock, node A measures the different time with B at identical some place fiducial time.By using different time variable t among Fig. 1 and Fig. 2
AAnd t
BCan clearly be seen that this point.Therefore, if at the particular point in time place, the clock deviation of the clock of Node B and node A side-play amount K, this deviation affects T
R (B)Measurement.Subsequently, if use T
S (A), T
R (B)And T
RR (A)Come computation migration to postpone or one-way latency (OWD), then it and actual or true OWD difference K mutually.In order to use this Cristian algorithm, must suppose that the true OWD on the forward direction equals the true OWD of back on direction, this is opposite with measured OWD.
Below, the time offset of being calculated between node A and Node B is expressed as Δ.This means that Δ is the estimation of K.In addition, the one-way latency of measuring on the forward and backward direction is expressed as OWD respectively
fAnd OWD
bCan calculate OWD according to following equation (1) and (2)
fAnd OWD
b
OWD
f＝T
R(B)-T
S(A) (1)
OWD
b＝T
RR(A)-T
R(B) (2)
According to the Cristian algorithm, can shown in following equation (3), calculate Δ.
Δ
CA＝(OWD
f-OWD
b)/2 (3)
Only notice when the forward direction that satisfies symmetry and back during Δ to the hypothesis of one-way latency
CAJust equal K.Yet, in real forward direction and back under the situation of the unequal asymmetric delays of OWD, the result that this estimation will lead to errors.As mentioned above, OWD is normally asymmetrical in live network.Therefore for real-time application, much more inessential than the error that causes by variable OWD by the systematic error that the asymmetrical network topology causes.Owing to can not provide the assurance to symmetrical one-way latency in live network, so the time causality (that is, node A sends grouping, and Node B receives this grouping and sends answers grouping) of three incidents only can be provided, this causes inequality T
S (A)＜T
R (B)＜T
RR (A)Therefore, if with T
R (B)Be estimated as at time point T
RR (A)And T
S (A)Between the centre, then max calculation error or the error boundary δ according to the estimated synchronisation offset of Cristian algorithm will satisfy following equation (4).
δ＝(T
RR(A)-T
S(A))/2＝RTT/2 (4)
In order to carry out the measurement of following method, advantageously satisfy some prerequisites.At first, network topology should not change in measuring period.Secondly, the time drift that causes owing to the difference of the clock frequency of the clock of node A and Node B should be very little, makes not change in the time difference between the internal clock during measuring period.If be several seconds magnitude measuring period, then almost satisfy these prerequisites forever.
For time synchronized, carry out measurement repeatedly according to Fig. 2.Fig. 3 shows in order to carry out this measurement sends a series of requests and answers the sequential chart of grouping between node A and Node B example.For the needs that illustrate, four communication cycles have only been described, each cycle comprises: divide into groups to Node B transmission request from node A, and answer grouping from Node B to node A transmission.The request grouping that is sent by node A is in time T as shown in Figure 3
S (A) 1, T
S (A) i, T
S (A) kAnd T
S (A) 1Send.Though the one-way latency that can calculate each communication cycle according to top equation (1) and (2), two-way time (RTT) can be calculated as corresponding OWD's and, perhaps alternatively and more simply be calculated as RTT=T
RR (A) i-T
S (A) iThe number of the communication cycle of measuring in fact can be bigger, a for example hundreds of measurement.
Between node A and Node B, exchange and ask and the answer grouping, till the approaching RTT of minimum minRTT two-way time of definite and measurement gathers.This set is called minRTT stability region on every side.For this reason, at first must select to be suitable for the maximum deviation d of given network topology
MaxAnd Minimum Area width W
MinStability region around the minRTT means the W at least that comprises the sampling with the RTT that equals minRTT
MinIndividual continuous sampling has to depart from minRTT and is less than or equal to d
MaxThe RTT value.
Can find the stability region by following process.At first, in initialization procedure, minRTT is set to infinity.Then, for each RTT communication cycle or request with to answer grouping right, calculate RTT and this RTT is compared with the minRTT value of being stored.If measured RTT is less than the minRTT that is stored, then minRTT is set to measured RTT.In addition, record departs from less than d with current minRTT
MaxThe number of measured RTT value.Depart from less than d in case write down with minRTT
MaxW at least
MinIndividual continuous RTT just logs on as these RTT values and stablizes the RTT value set.Also login corresponding one-way latency and in further calculating, use the estimation of this corresponding one-way latency with the synchronisation offset between the clock that obtains node A and Node B.
In next procedure, calculate minimal forward one-way latency (minF) and minimum back to one-way latency (minB) according to following equation (5) and (6).Thereby the minimal forward one-way latency is the minimum value of login in forward direction one-way latency set, and minimum back is to login the minimum value in the one-way latency set in the back to one-way latency.Along with the number of data packets purpose growth that exchanges between node A and the Node B, packet increases through the probability of waiting list, and wherein said waiting list is empty fully and the processing time cause minimum at intranodal.Therefore the physics limit in the one-way latency is represented in the minimum one-way latency on might forward and backward direction.Physics limit means that the element such as waiting list does not influence delay.
minF＝min(T
R(B)j-T
S(A)j)，1＜＝j＜＝W
min(5)
minB＝min(T
RR(A)n-T
R(B)n)，1＜＝n＜＝W
min(6)
Utilize following equation (7) to calculate so-called virtual minimum RTT (VirtMinRTT) then.
VirtMinRTT＝minF+minB (7)
Therefore, suppose minF and minB be the packet minimum that may experience, possible one-way latency physically, VirtMinRTT is at being sent to Node B from node A and defining the minimum two-way time that is subjected to the physical boundary restriction from the grouping that Node B sends it back node A.
Fig. 5 shows from node A and is sent to Node B and sends it back the minRTT of series of data packets of node A and the figure of VirtMinRTT from Node B.Herein, abscissa shows the number of measurement.Use the process of above-mentioned searching stability region and calculating minRTT and VirtMinRTT to carry out measurement each time.Usually, VirtMinRTT is not equal to minRTT.This means the grouping of minimum queuing of the forward path from node A to Node B experience or processing delay be not same grouping in back grouping to the path experience minimum delay from Node B to node A.As shown in Figure 5, VirtMinRTT is lower than minRTT all the time.
Owing to the constant network topology on measuring period is assumed to precondition, so minF and minB equal or be approximately equal to the bounded above minimum one-way latency of physics at least.Yet, because the clock of the clock of Node B and node A is asynchronous, so minF and minB represent measured one-way latency, rather than true or actual delay.If the clock of Node B has unknown clock synchronization side-play amount K with respect to the clock of node A, then can use following equation (8) and (9) to calculate real one-way latency minF
rAnd minB
rThis means minF
rAnd minB
rRespectively the actual minimum one-way latency on the forward and backward direction when correcting by synchronisation offset K.
minF
r＝minF-K (8)
minB
r＝minB+K (9)
Can be with equation (8) and (9) substitution equation (7) obtaining following equation (10), this equation shows the value with the irrelevant VirtMinRTT of synchronisation offset.
VirtMinRTT＝minF+minB＝minF
r+minB
r (10)
Utilize with the corresponding forward direction of VirtMinRTT and back to one-way latency minF and minB, the estimation that can calculate synchronisation offset according to equation (11).
Δ＝(minF-minB)/2 (11)
Fig. 6 shows the figure of the calculated value (dotted line) of Δ.Can provide estimation approach to be called improved Cristian algorithm (ICA) with this to synchronisation offset.This algorithm is different significantly with above-described Cristian algorithm (CA).At first, based on a series of measurements, and CA is based on single measurement according to the method for ICA.In addition, according to the calculating of ICA based on determining forward direction and back minimum value minF and minB independently to one-way latency, and among the CA employed forward direction and back to one-way latency OWD
fAnd OWD
bBe to send it back the value that node A determines from the B node same communication cycle by sending grouping and will divide into groups to Node B from node A.As a result of, Δ compares Δ
CAHave more consistency, and therefore provide estimation more accurately, as shown in Figure 6, in Fig. 6, drawn the Δ that calculates at identical measuring assembly synchronisation offset K
CAFor comparing.
As alternative, can use with the corresponding forward direction of minimum RTT two-way time and back to one-way latency, rather than use with as the corresponding one-way latency of the VirtMinRTT among the ICA, calculate the difference estimation Δ of synchronisation offset
CAminThis method will be called the CAmin method.In Fig. 7, drawn this result of calculation at measuring assembly, and the Δ that uses ICA to calculate at this measuring assembly.As shown in Figure 7, Δ has all the time and is lower than (meaning more the mean value near Δ) or equals Δ
CAminBorder (dotted line).
The calculating of Δ is measured leg-of-mutton centre based on estimation is placed on.Therefore, as shown in Figure 4, the certainty border δ of the error of calculation
1Based on postpone T
S (A), T
R (B)And T
RR (A)The time causality of relevant incident.Can use equation (12) to calculate certainty border δ
1
δ
1＝VirtMinRTT/2 (12)
For VirtMinRTT, minRTT and RTT
i(represent any single two-way time value), following inequality (13) is set up.
VirtMinRTT＜＝minRTT＜＝RTT
i (13)
This inequality is hinting error boundary δ
1The same at least good with the error boundary that uses traditional C ristian algorithm to estimate.In addition, if VirtMinRTT is not equal to minRTT, then compare acquisition error boundary more closely with using any other measurement point.Owing to this conclusion certainty ground can be applied to the algorithm that proposed, so with δ
1Be called the ascertainment error border.
Fig. 8 shows the error boundary δ that estimates according to the side-play amount of traditional C ristian algorithm
CAFigure, Fig. 9 shows the error boundary δ of the estimation that obtains according to above-mentioned CAmin method and ICA method respectively
CAmin(thick dashed line) and δ
ICAThe figure of (fine dotted line).Can see that by comparison diagram 8 and Fig. 9 the Cristian algorithm causes high level error boundary graph, and the ICA method causes the minimum error boundary graph, thereby the most accurate estimation to synchronisation offset is provided.
From angle at random, can be with the time range combination of incident tightr.If minF and minB represent the hypothesis of the physical boundary of one-way latency and set up, then can define error boundary as following equation (14), it equals half of difference of minRTT and VirtMinRTT.Then this error boundary is called statistical error bound.
δ
2＝(minRTT-VirtMinRTT)/2(14)
Figure 10 shows the figure of the statistical error bound of being calculated.Statistical error bound has following meaning.Estimated synchronisation offset Δ is at the statistical error bound δ of distance real synchronisation offset K
2In the probability value of being greater than or equal to minF and minB be the probability of the bounded above one-way latency of physics really.In other words, even minF and minB are not the minimum physics possibility one-way latencys of transmitted in packets, δ
2For the error boundary of estimated synchronisation offset Δ, also be useful numeral.Can see by comparison diagram 9 and Figure 10, with ascertainment error border δ
1That is advised compares, according to statistical error bound δ
2, estimated synchronisation offset Δ can be the better estimation to K.
In the above description, the feature of the present invention that discloses in the claim and in the accompanying drawing is for separately and to implement different embodiments of the invention with the form of any required combination be important.
Claims (8)
1, a kind of clock of networking component in the network (A) and method of the clock synchronization of the other networking component (B) that communicates to connect with this networking component (A) of making, described method comprises the following steps:
-be identified for from the forward transition delay set of described networking component (A) to described other networking component (B) message transfer, and be used for from described other networking component (B) back to the transition delay set to described networking component (A) message transfer,
-from the set of forward direction transition delay, select minimum forward transition delay (minF),
-to the transition delay set, select minimum back to transition delay (minB) from the back,
-from selected minimum forward transition delay (minF) and minimum back to transition delay (minB) the estimated value of the side-play amount of derivation between the clock of the clock of described networking component (A) and described other networking component (B), and
-described the estimated value of estimating according to side-play amount is adjusted the clock of described networking component (A).
2, method according to claim 1, it is characterized in that, determine that forward transition delay set and back also comprise step to the step that transition delay is gathered: be identified for continuous two-way time of between described networking component (A) and described other networking component (B), communicating by letter, come down to corresponding forward transition delay and corresponding back wherein said two-way time to the transition delay sum, comprise predetermined number up to finding, till departing from less than predetermined maximum deviation (dmax) with minimum two-way time (minRTT) two-way time of the two-way time of Que Dinging in interior stable set two-way time makes described stable set two-way time continuously, will with the corresponding forward transition delay login two-way time in stable two-way time of the set in the forward transition delay set, and will with in stable set two-way time two-way time corresponding back to the transition delay login in the back in the transition delay set.
3, method according to claim 1 and 2 is characterized in that, determines that forward transition delay and corresponding back comprise to the step of transition delay: send request from described networking component (A) to described other networking component (B) at request time (TS); Send answer to described request from described other networking component (B) to described networking component (A) in the time for replying (TR); Receive described answer at the answer described networking component of time of reception (TRR) (A), calculate described forward transition delay by from the described time for replying (TR), deducting the described request time (TS); And calculate corresponding back to transition delay by from described answer time of reception (TRR), deducting the described time for replying (TR).
4, method according to claim 3 also is included in the step that encapsulates timestamp from described other networking component (B) to each answer that described networking component (A) sends, and wherein said timestamp comprises the time for replying (TR).
5, according to claim 3 or 4 described methods, also comprise the step that sends described request with equal time at interval.
6, according to any described method in the aforementioned claim 3 to 5, it is characterized in that described request and described answer send as packet.
7, according to any described method in the aforementioned claim, it is characterized in that, upgrade forward transition delay set and back continuously and gather to transition delay.
8, be used for putting into the networking component (A) of network, comprise:
-clock,
-communicator, be configured to provide with network in being connected of at least one other networking component (B),
-request unit is connected with described communicator, and be configured to generate request and transmit described request to described other networking component (B),
-receiving system is connected with described communicator, is configured to receive answer to described request from described other networking component (B),
-analytical equipment is connected with described clock with described receiving system, is configured to the answer that receives is analyzed, determined transition delay set and calculates estimation to clock offset,
-memory is connected with described analytical equipment, is configured to store the transition delay set, and
-clock adjusting device is connected with described analytical equipment, is configured to estimate to adjust described clock according to institute's offset calculated.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| EP07107929.7 | 2007-05-10 | ||
| EP07107929A EP1990938A1 (en) | 2007-05-10 | 2007-05-10 | Method for synchronizing a clock of a network component with a clock of a further network component and network component therefor |
| PCT/EP2008/055677 WO2008138850A1 (en) | 2007-05-10 | 2008-05-08 | Method for synchronizing a clock of a network component with a clock of further network component and network component therefor |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101675614A true CN101675614A (en) | 2010-03-17 |
Family
ID=38289931
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200880013892A Pending CN101675614A (en) | 2007-05-10 | 2008-05-08 | Method for synchronizing a clock of a network component with a clock of further network component and network component therefor |
Country Status (6)
| Country | Link |
|---|---|
| US (1) | US20100135332A1 (en) |
| EP (2) | EP1990938A1 (en) |
| JP (1) | JP2010527193A (en) |
| KR (1) | KR20100016364A (en) |
| CN (1) | CN101675614A (en) |
| WO (1) | WO2008138850A1 (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102255791A (en) * | 2010-05-21 | 2011-11-23 | 中国移动通信集团公司 | Inter-equipment link delay determination method and system |
| CN104601393A (en) * | 2013-10-30 | 2015-05-06 | 三星Sds株式会社 | Method and apparatus for estimating queuing delay |
| CN104813303A (en) * | 2012-09-14 | 2015-07-29 | Dts（英属维尔京群岛）有限公司 | Playback synchronization |
| CN105594146A (en) * | 2013-10-02 | 2016-05-18 | 哈利法科学技术研究大学 | Method and devices for compensating for path asymmetry |
| CN107251456A (en) * | 2014-12-24 | 2017-10-13 | Abb瑞士股份有限公司 | Method for carrying out time synchronized to the device in control network |
| CN108474834A (en) * | 2016-01-06 | 2018-08-31 | 阿尔卡特朗讯美国公司 | The method and apparatus synchronized for aerial anchor point-anchor point |
| CN109716683A (en) * | 2016-09-30 | 2019-05-03 | 网络洞察力知识产权公司 | Time synchronization in real time content dissemination system |
Families Citing this family (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8139489B2 (en) | 2008-08-06 | 2012-03-20 | International Business Machines Corporation | Robust jitter-free remote clock offset measuring method |
| KR101114702B1 (en) | 2008-12-18 | 2012-02-29 | 삼성전자주식회사 | Method and system for time synchronization among nodes in a wireless communication system |
| KR101182861B1 (en) * | 2008-12-22 | 2012-09-14 | 한국전자통신연구원 | Distance measuring method |
| US8108557B2 (en) * | 2009-01-22 | 2012-01-31 | Hewlett-Packard Development Company, L.P. | System and method for measuring clock skew on a network |
| JP5448679B2 (en) * | 2009-09-25 | 2014-03-19 | 三菱電機株式会社 | Wireless base station equipment |
| US8533355B2 (en) | 2009-11-02 | 2013-09-10 | International Business Machines Corporation | Techniques for improved clock offset measuring |
| KR101232929B1 (en) | 2009-11-23 | 2013-02-15 | 한국전자통신연구원 | Method and apparatus for time synchronization of distributed nodes over wideband high frequency wireless |
| DE102010049121B8 (en) | 2010-10-21 | 2014-06-18 | Hochschule Anhalt (Fh) | Method for illuminating roads, paths, squares or rooms and communication unit for use in the method |
| JP5741010B2 (en) * | 2011-01-26 | 2015-07-01 | 日本電気株式会社 | Synchronization system |
| EP2627040B1 (en) | 2012-02-12 | 2018-02-28 | Hochschule Anhalt (FH) | Method for eliminating systematical error components in a set of one-way delay measurement results for communications between two or more computing systems in a communication network, apparatus for performing the method and computer program product |
| JP5675703B2 (en) * | 2012-06-04 | 2015-02-25 | 株式会社東芝 | Communication device, control device, and program |
| US9247448B2 (en) | 2012-08-27 | 2016-01-26 | Qualcomm Incorporated | Device and method for adaptive rate multimedia communications on a wireless network |
| US10051519B2 (en) | 2012-08-27 | 2018-08-14 | Qualcomm Incorporated | Device and method for adaptive rate multimedia communications on a wireless network |
| JP6010802B2 (en) * | 2013-02-22 | 2016-10-19 | 東日本電信電話株式会社 | Time synchronization system, time synchronization method, slave node, and computer program |
| JP2015039131A (en) * | 2013-08-19 | 2015-02-26 | 株式会社東芝 | Measurement device and method |
| US9112628B2 (en) * | 2013-10-02 | 2015-08-18 | Khalifa University of Science, Technology, and Research | Method and devices for compensating for path asymmetry |
| WO2015116980A1 (en) * | 2014-01-31 | 2015-08-06 | University Of North Dakota | Network clock skew estimation and calibration |
| JP6501138B2 (en) * | 2014-07-18 | 2019-04-17 | セイコーソリューションズ株式会社 | Delay measurement method, delay measurement device, and program |
| JP6523058B2 (en) * | 2015-06-05 | 2019-05-29 | 株式会社東芝 | Communication system and estimation method |
| EP3545634B1 (en) * | 2016-11-22 | 2024-08-07 | Khalifa University of Science and Technology | Methods and systems for determining optimal messages |
| DE102018208210A1 (en) * | 2018-05-24 | 2019-11-28 | Siemens Aktiengesellschaft | Apparatus and method for determining a time offset between two time bases |
| CN111556559B (en) * | 2020-05-09 | 2021-11-26 | 重庆邮电大学 | Hybrid clock synchronization method based on timestamp-free interaction and one-way message propagation |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3204609B2 (en) * | 1995-12-14 | 2001-09-04 | 松下電器産業株式会社 | Cell receiver having source clock recovery circuit |
| JP3480701B2 (en) * | 1999-08-06 | 2003-12-22 | 日本電気株式会社 | Packet network transmission delay measuring method and machine-readable recording medium recording program |
| JP2001111623A (en) * | 1999-10-14 | 2001-04-20 | Nippon Telegr & Teleph Corp <Ntt> | Distributed time synchronization method and distributed time synchronization system utilizing this method |
| US6438702B1 (en) * | 1999-12-21 | 2002-08-20 | Telcordia Technologies, Inc. | Method for providing a precise network time service |
| EP1209850B1 (en) * | 2000-11-22 | 2007-04-04 | Lucent Technologies Inc. | System and method for network element synchronization |
| US7765422B2 (en) * | 2001-01-19 | 2010-07-27 | Alcatel-Lucent Usa Inc. | Method of determining a time offset estimate between a central node and a secondary node |
| US7194556B2 (en) * | 2001-03-30 | 2007-03-20 | Intel Corporation | Method and apparatus for high accuracy distributed time synchronization using processor tick counters |
| DE60213623T2 (en) * | 2002-12-09 | 2007-10-18 | Tektronix International Sales Gmbh | Circulation time estimation method and means by acknowledgment in a packet transmission system |
| US7688865B2 (en) * | 2005-09-09 | 2010-03-30 | International Business Machines Corporation | Method and system for clock skew and offset estimation |
| US7822072B2 (en) * | 2005-09-09 | 2010-10-26 | International Business Machines Corporation | Clock filter dispersion |
| KR100813966B1 (en) * | 2005-11-22 | 2008-03-14 | 연세대학교 산학협력단 | Clock synchronization method and apparatus using estimated one-way delay |
-
2007
- 2007-05-10 EP EP07107929A patent/EP1990938A1/en not_active Withdrawn
-
2008
- 2008-05-08 US US12/451,364 patent/US20100135332A1/en not_active Abandoned
- 2008-05-08 JP JP2010506936A patent/JP2010527193A/en active Pending
- 2008-05-08 CN CN200880013892A patent/CN101675614A/en active Pending
- 2008-05-08 WO PCT/EP2008/055677 patent/WO2008138850A1/en active Application Filing
- 2008-05-08 KR KR1020097023368A patent/KR20100016364A/en not_active Application Discontinuation
- 2008-05-08 EP EP08759462A patent/EP2145412A1/en not_active Withdrawn
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102255791B (en) * | 2010-05-21 | 2013-12-18 | 中国移动通信集团公司 | Inter-equipment link delay determination method and system |
| CN102255791A (en) * | 2010-05-21 | 2011-11-23 | 中国移动通信集团公司 | Inter-equipment link delay determination method and system |
| CN104813303A (en) * | 2012-09-14 | 2015-07-29 | Dts（英属维尔京群岛）有限公司 | Playback synchronization |
| CN104813303B (en) * | 2012-09-14 | 2018-11-16 | Dts（英属维尔京群岛）有限公司 | Playback synchronizes |
| CN105594146A (en) * | 2013-10-02 | 2016-05-18 | 哈利法科学技术研究大学 | Method and devices for compensating for path asymmetry |
| CN105594146B (en) * | 2013-10-02 | 2018-04-27 | 哈利法科学技术研究大学 | For compensating the asymmetric method and apparatus in path |
| CN104601393A (en) * | 2013-10-30 | 2015-05-06 | 三星Sds株式会社 | Method and apparatus for estimating queuing delay |
| CN104601393B (en) * | 2013-10-30 | 2018-05-29 | 三星Sds株式会社 | Queueing delay method of estimation and device |
| CN107251456B (en) * | 2014-12-24 | 2019-03-29 | Abb瑞士股份有限公司 | Method for carrying out time synchronization to the device in control network |
| CN107251456A (en) * | 2014-12-24 | 2017-10-13 | Abb瑞士股份有限公司 | Method for carrying out time synchronized to the device in control network |
| CN108474834A (en) * | 2016-01-06 | 2018-08-31 | 阿尔卡特朗讯美国公司 | The method and apparatus synchronized for aerial anchor point-anchor point |
| CN109716683A (en) * | 2016-09-30 | 2019-05-03 | 网络洞察力知识产权公司 | Time synchronization in real time content dissemination system |
| CN109716683B (en) * | 2016-09-30 | 2021-03-19 | 网络洞察力有限公司 | Time synchronization in real-time content distribution systems |
| US10972775B2 (en) | 2016-09-30 | 2021-04-06 | Net Insight Ab | Time synchronization of client devices using two-way time transfer in a live content distribution system |
| CN112866758A (en) * | 2016-09-30 | 2021-05-28 | 网络洞察力有限公司 | Time synchronization in real-time content distribution systems |
| US11659215B2 (en) | 2016-09-30 | 2023-05-23 | Net Insight Ab | Time synchronization in a live content distribution system |
| CN112866758B (en) * | 2016-09-30 | 2024-01-05 | 网络洞察力有限公司 | Method for use in a network system capable of node-to-node time transfer |
Also Published As
| Publication number | Publication date |
|---|---|
| EP2145412A1 (en) | 2010-01-20 |
| KR20100016364A (en) | 2010-02-12 |
| WO2008138850A1 (en) | 2008-11-20 |
| JP2010527193A (en) | 2010-08-05 |
| US20100135332A1 (en) | 2010-06-03 |
| EP1990938A1 (en) | 2008-11-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101675614A (en) | Method for synchronizing a clock of a network component with a clock of further network component and network component therefor | |
| US7876791B2 (en) | Synchronizing apparatus and method in packet network | |
| US10615898B2 (en) | System and method of clock management in a packet data network | |
| CN111585683B (en) | High-reliability clock synchronization system and method for time-sensitive network | |
| US8416812B2 (en) | Network timing synchronization systems | |
| EP2342850B1 (en) | A method for synchronizing clocks in a communication network | |
| De Vito et al. | One-way delay measurement: State of the art | |
| US10396971B2 (en) | One-way packet delay measurement | |
| US8018972B2 (en) | Timing over packet performance | |
| EP2374245B1 (en) | Controlling packet transmission using bandwidth estimation | |
| CN102197611B (en) | Method and device for packet network synchronization | |
| EP2410672B1 (en) | Method and system for transmitting time in passive optical network | |
| US9270607B2 (en) | Method and devices for packet selection | |
| EP1961186B1 (en) | Using travel-time as means for improving the accuracy of simple network time protocol | |
| US11153838B1 (en) | Time synchronization enhancements for integrated 5G-TSN networking | |
| KR20090024170A (en) | Network time protocol precision timestamping service | |
| CN101631013B (en) | Method, equipment and system for clock synchronization of access network | |
| CN103929293A (en) | Asymmetrically-delayed time synchronization method and system | |
| Shin et al. | Clock synchronization for one-way delay measurement: A survey | |
| US6647026B1 (en) | Frame phase synchronous system and a method thereof | |
| EP2627040B1 (en) | Method for eliminating systematical error components in a set of one-way delay measurement results for communications between two or more computing systems in a communication network, apparatus for performing the method and computer program product | |
| KR101315593B1 (en) | Method for synchronizing a client clock frequency with a server clock frequency | |
| KR20090120998A (en) | Time synchronizing method and apparatus of network | |
| EP3080951B1 (en) | Method and devices for packet selection | |
| Kim et al. | One-way delay estimation without clock sychronization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| ASS | Succession or assignment of patent right |
Owner name: DIESEL CO., LTD. Free format text: FORMER OWNER: THOMSON LICENSING TRADE CO. Effective date: 20110316 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: BOULOGNE-BILLANCOURT, FRANCE TO: HANNOVER, GERMANY |
|
| TA01 | Transfer of patent application right |
Effective date of registration: 20110316 Address after: Hannover Applicant after: Tixel GmbH Address before: French Boulogne - Bilang Kurt Applicant before: Thomson Licensing Trade Co. |
|
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20100317 |